diff options
Diffstat (limited to 'arch')
36 files changed, 412 insertions, 1336 deletions
diff --git a/arch/alpha/Kconfig b/arch/alpha/Kconfig index d5b9b5e645cc..9944dedee5b1 100644 --- a/arch/alpha/Kconfig +++ b/arch/alpha/Kconfig @@ -18,6 +18,8 @@ config ALPHA select ARCH_HAVE_NMI_SAFE_CMPXCHG select GENERIC_SMP_IDLE_THREAD select GENERIC_CMOS_UPDATE + select GENERIC_STRNCPY_FROM_USER + select GENERIC_STRNLEN_USER help The Alpha is a 64-bit general-purpose processor designed and marketed by the Digital Equipment Corporation of blessed memory, diff --git a/arch/alpha/include/asm/atomic.h b/arch/alpha/include/asm/atomic.h index 3bb7ffeae3bc..c2cbe4fc391c 100644 --- a/arch/alpha/include/asm/atomic.h +++ b/arch/alpha/include/asm/atomic.h @@ -14,8 +14,8 @@ */ -#define ATOMIC_INIT(i) ( (atomic_t) { (i) } ) -#define ATOMIC64_INIT(i) ( (atomic64_t) { (i) } ) +#define ATOMIC_INIT(i) { (i) } +#define ATOMIC64_INIT(i) { (i) } #define atomic_read(v) (*(volatile int *)&(v)->counter) #define atomic64_read(v) (*(volatile long *)&(v)->counter) diff --git a/arch/alpha/include/asm/fpu.h b/arch/alpha/include/asm/fpu.h index db00f7885faa..e477bcd5b94a 100644 --- a/arch/alpha/include/asm/fpu.h +++ b/arch/alpha/include/asm/fpu.h @@ -1,7 +1,9 @@ #ifndef __ASM_ALPHA_FPU_H #define __ASM_ALPHA_FPU_H +#ifdef __KERNEL__ #include <asm/special_insns.h> +#endif /* * Alpha floating-point control register defines: diff --git a/arch/alpha/include/asm/ptrace.h b/arch/alpha/include/asm/ptrace.h index fd698a174f26..b87755a19554 100644 --- a/arch/alpha/include/asm/ptrace.h +++ b/arch/alpha/include/asm/ptrace.h @@ -76,7 +76,10 @@ struct switch_stack { #define task_pt_regs(task) \ ((struct pt_regs *) (task_stack_page(task) + 2*PAGE_SIZE) - 1) -#define force_successful_syscall_return() (task_pt_regs(current)->r0 = 0) +#define current_pt_regs() \ + ((struct pt_regs *) ((char *)current_thread_info() + 2*PAGE_SIZE) - 1) + +#define force_successful_syscall_return() (current_pt_regs()->r0 = 0) #endif diff --git a/arch/alpha/include/asm/socket.h b/arch/alpha/include/asm/socket.h index dcb221a4b5be..7d2f75be932e 100644 --- a/arch/alpha/include/asm/socket.h +++ b/arch/alpha/include/asm/socket.h @@ -76,9 +76,11 @@ /* Instruct lower device to use last 4-bytes of skb data as FCS */ #define SO_NOFCS 43 +#ifdef __KERNEL__ /* O_NONBLOCK clashes with the bits used for socket types. Therefore we * have to define SOCK_NONBLOCK to a different value here. */ #define SOCK_NONBLOCK 0x40000000 +#endif /* __KERNEL__ */ #endif /* _ASM_SOCKET_H */ diff --git a/arch/alpha/include/asm/uaccess.h b/arch/alpha/include/asm/uaccess.h index b49ec2f8d6e3..766fdfde2b7a 100644 --- a/arch/alpha/include/asm/uaccess.h +++ b/arch/alpha/include/asm/uaccess.h @@ -433,36 +433,12 @@ clear_user(void __user *to, long len) #undef __module_address #undef __module_call -/* Returns: -EFAULT if exception before terminator, N if the entire - buffer filled, else strlen. */ +#define user_addr_max() \ + (segment_eq(get_fs(), USER_DS) ? TASK_SIZE : ~0UL) -extern long __strncpy_from_user(char *__to, const char __user *__from, long __to_len); - -extern inline long -strncpy_from_user(char *to, const char __user *from, long n) -{ - long ret = -EFAULT; - if (__access_ok((unsigned long)from, 0, get_fs())) - ret = __strncpy_from_user(to, from, n); - return ret; -} - -/* Returns: 0 if bad, string length+1 (memory size) of string if ok */ -extern long __strlen_user(const char __user *); - -extern inline long strlen_user(const char __user *str) -{ - return access_ok(VERIFY_READ,str,0) ? __strlen_user(str) : 0; -} - -/* Returns: 0 if exception before NUL or reaching the supplied limit (N), - * a value greater than N if the limit would be exceeded, else strlen. */ -extern long __strnlen_user(const char __user *, long); - -extern inline long strnlen_user(const char __user *str, long n) -{ - return access_ok(VERIFY_READ,str,0) ? __strnlen_user(str, n) : 0; -} +extern long strncpy_from_user(char *dest, const char __user *src, long count); +extern __must_check long strlen_user(const char __user *str); +extern __must_check long strnlen_user(const char __user *str, long n); /* * About the exception table: diff --git a/arch/alpha/include/asm/unistd.h b/arch/alpha/include/asm/unistd.h index 633b23b0664a..a31a78eac9b9 100644 --- a/arch/alpha/include/asm/unistd.h +++ b/arch/alpha/include/asm/unistd.h @@ -465,10 +465,12 @@ #define __NR_setns 501 #define __NR_accept4 502 #define __NR_sendmmsg 503 +#define __NR_process_vm_readv 504 +#define __NR_process_vm_writev 505 #ifdef __KERNEL__ -#define NR_SYSCALLS 504 +#define NR_SYSCALLS 506 #define __ARCH_WANT_OLD_READDIR #define __ARCH_WANT_STAT64 diff --git a/arch/alpha/include/asm/word-at-a-time.h b/arch/alpha/include/asm/word-at-a-time.h new file mode 100644 index 000000000000..6b340d0f1521 --- /dev/null +++ b/arch/alpha/include/asm/word-at-a-time.h @@ -0,0 +1,55 @@ +#ifndef _ASM_WORD_AT_A_TIME_H +#define _ASM_WORD_AT_A_TIME_H + +#include <asm/compiler.h> + +/* + * word-at-a-time interface for Alpha. + */ + +/* + * We do not use the word_at_a_time struct on Alpha, but it needs to be + * implemented to humour the generic code. + */ +struct word_at_a_time { + const unsigned long unused; +}; + +#define WORD_AT_A_TIME_CONSTANTS { 0 } + +/* Return nonzero if val has a zero */ +static inline unsigned long has_zero(unsigned long val, unsigned long *bits, const struct word_at_a_time *c) +{ + unsigned long zero_locations = __kernel_cmpbge(0, val); + *bits = zero_locations; + return zero_locations; +} + +static inline unsigned long prep_zero_mask(unsigned long val, unsigned long bits, const struct word_at_a_time *c) +{ + return bits; +} + +#define create_zero_mask(bits) (bits) + +static inline unsigned long find_zero(unsigned long bits) +{ +#if defined(CONFIG_ALPHA_EV6) && defined(CONFIG_ALPHA_EV67) + /* Simple if have CIX instructions */ + return __kernel_cttz(bits); +#else + unsigned long t1, t2, t3; + /* Retain lowest set bit only */ + bits &= -bits; + /* Binary search for lowest set bit */ + t1 = bits & 0xf0; + t2 = bits & 0xcc; + t3 = bits & 0xaa; + if (t1) t1 = 4; + if (t2) t2 = 2; + if (t3) t3 = 1; + return t1 + t2 + t3; +#endif +} + +#endif /* _ASM_WORD_AT_A_TIME_H */ diff --git a/arch/alpha/kernel/alpha_ksyms.c b/arch/alpha/kernel/alpha_ksyms.c index d96e742d4dc2..15fa821d09cd 100644 --- a/arch/alpha/kernel/alpha_ksyms.c +++ b/arch/alpha/kernel/alpha_ksyms.c @@ -52,7 +52,6 @@ EXPORT_SYMBOL(alpha_write_fp_reg_s); /* entry.S */ EXPORT_SYMBOL(kernel_thread); -EXPORT_SYMBOL(kernel_execve); /* Networking helper routines. */ EXPORT_SYMBOL(csum_tcpudp_magic); @@ -74,8 +73,6 @@ EXPORT_SYMBOL(alpha_fp_emul); */ EXPORT_SYMBOL(__copy_user); EXPORT_SYMBOL(__do_clear_user); -EXPORT_SYMBOL(__strncpy_from_user); -EXPORT_SYMBOL(__strnlen_user); /* * SMP-specific symbols. diff --git a/arch/alpha/kernel/entry.S b/arch/alpha/kernel/entry.S index 6d159cee5f2f..ec0da0567ab5 100644 --- a/arch/alpha/kernel/entry.S +++ b/arch/alpha/kernel/entry.S @@ -663,58 +663,6 @@ kernel_thread: br ret_to_kernel .end kernel_thread -/* - * kernel_execve(path, argv, envp) - */ - .align 4 - .globl kernel_execve - .ent kernel_execve -kernel_execve: - /* We can be called from a module. */ - ldgp $gp, 0($27) - lda $sp, -(32+SIZEOF_PT_REGS+8)($sp) - .frame $sp, 32+SIZEOF_PT_REGS+8, $26, 0 - stq $26, 0($sp) - stq $16, 8($sp) - stq $17, 16($sp) - stq $18, 24($sp) - .prologue 1 - - lda $16, 32($sp) - lda $17, 0 - lda $18, SIZEOF_PT_REGS - bsr $26, memset !samegp - - /* Avoid the HAE being gratuitously wrong, which would cause us - to do the whole turn off interrupts thing and restore it. */ - ldq $2, alpha_mv+HAE_CACHE - stq $2, 152+32($sp) - - ldq $16, 8($sp) - ldq $17, 16($sp) - ldq $18, 24($sp) - lda $19, 32($sp) - bsr $26, do_execve !samegp - - ldq $26, 0($sp) - bne $0, 1f /* error! */ - - /* Move the temporary pt_regs struct from its current location - to the top of the kernel stack frame. See copy_thread for - details for a normal process. */ - lda $16, 0x4000 - SIZEOF_PT_REGS($8) - lda $17, 32($sp) - lda $18, SIZEOF_PT_REGS - bsr $26, memmove !samegp - - /* Take that over as our new stack frame and visit userland! */ - lda $sp, 0x4000 - SIZEOF_PT_REGS($8) - br $31, ret_from_sys_call - -1: lda $sp, 32+SIZEOF_PT_REGS+8($sp) - ret -.end kernel_execve - /* * Special system calls. Most of these are special in that they either @@ -797,115 +745,6 @@ sys_rt_sigreturn: .end sys_rt_sigreturn .align 4 - .globl sys_sethae - .ent sys_sethae -sys_sethae: - .prologue 0 - stq $16, 152($sp) - ret -.end sys_sethae - - .align 4 - .globl osf_getpriority - .ent osf_getpriority -osf_getpriority: - lda $sp, -16($sp) - stq $26, 0($sp) - .prologue 0 - - jsr $26, sys_getpriority - - ldq $26, 0($sp) - blt $0, 1f - - /* Return value is the unbiased priority, i.e. 20 - prio. - This does result in negative return values, so signal - no error by writing into the R0 slot. */ - lda $1, 20 - stq $31, 16($sp) - subl $1, $0, $0 - unop - -1: lda $sp, 16($sp) - ret -.end osf_getpriority - - .align 4 - .globl sys_getxuid - .ent sys_getxuid -sys_getxuid: - .prologue 0 - ldq $2, TI_TASK($8) - ldq $3, TASK_CRED($2) - ldl $0, CRED_UID($3) - ldl $1, CRED_EUID($3) - stq $1, 80($sp) - ret -.end sys_getxuid - - .align 4 - .globl sys_getxgid - .ent sys_getxgid -sys_getxgid: - .prologue 0 - ldq $2, TI_TASK($8) - ldq $3, TASK_CRED($2) - ldl $0, CRED_GID($3) - ldl $1, CRED_EGID($3) - stq $1, 80($sp) - ret -.end sys_getxgid - - .align 4 - .globl sys_getxpid - .ent sys_getxpid -sys_getxpid: - .prologue 0 - ldq $2, TI_TASK($8) - - /* See linux/kernel/timer.c sys_getppid for discussion - about this loop. */ - ldq $3, TASK_GROUP_LEADER($2) - ldq $4, TASK_REAL_PARENT($3) - ldl $0, TASK_TGID($2) -1: ldl $1, TASK_TGID($4) -#ifdef CONFIG_SMP - mov $4, $5 - mb - ldq $3, TASK_GROUP_LEADER($2) - ldq $4, TASK_REAL_PARENT($3) - cmpeq $4, $5, $5 - beq $5, 1b -#endif - stq $1, 80($sp) - ret -.end sys_getxpid - - .align 4 - .globl sys_alpha_pipe - .ent sys_alpha_pipe -sys_alpha_pipe: - lda $sp, -16($sp) - stq $26, 0($sp) - .prologue 0 - - mov $31, $17 - lda $16, 8($sp) - jsr $26, do_pipe_flags - - ldq $26, 0($sp) - bne $0, 1f - - /* The return values are in $0 and $20. */ - ldl $1, 12($sp) - ldl $0, 8($sp) - - stq $1, 80+16($sp) -1: lda $sp, 16($sp) - ret -.end sys_alpha_pipe - - .align 4 .globl sys_execve .ent sys_execve sys_execve: diff --git a/arch/alpha/kernel/osf_sys.c b/arch/alpha/kernel/osf_sys.c index 98a103621af6..bc1acdda7a5e 100644 --- a/arch/alpha/kernel/osf_sys.c +++ b/arch/alpha/kernel/osf_sys.c @@ -1404,3 +1404,52 @@ SYSCALL_DEFINE3(osf_writev, unsigned long, fd, } #endif + +SYSCALL_DEFINE2(osf_getpriority, int, which, int, who) +{ + int prio = sys_getpriority(which, who); + if (prio >= 0) { + /* Return value is the unbiased priority, i.e. 20 - prio. + This does result in negative return values, so signal + no error */ + force_successful_syscall_return(); + prio = 20 - prio; + } + return prio; +} + +SYSCALL_DEFINE0(getxuid) +{ + current_pt_regs()->r20 = sys_geteuid(); + return sys_getuid(); +} + +SYSCALL_DEFINE0(getxgid) +{ + current_pt_regs()->r20 = sys_getegid(); + return sys_getgid(); +} + +SYSCALL_DEFINE0(getxpid) +{ + current_pt_regs()->r20 = sys_getppid(); + return sys_getpid(); +} + +SYSCALL_DEFINE0(alpha_pipe) +{ + int fd[2]; + int res = do_pipe_flags(fd, 0); + if (!res) { + /* The return values are in $0 and $20. */ + current_pt_regs()->r20 = fd[1]; + res = fd[0]; + } + return res; +} + +SYSCALL_DEFINE1(sethae, unsigned long, val) +{ + current_pt_regs()->hae = val; + return 0; +} diff --git a/arch/alpha/kernel/process.c b/arch/alpha/kernel/process.c index 153d3fce3e8e..d6fde98b74b3 100644 --- a/arch/alpha/kernel/process.c +++ b/arch/alpha/kernel/process.c @@ -455,3 +455,22 @@ get_wchan(struct task_struct *p) } return pc; } + +int kernel_execve(const char *path, const char *const argv[], const char *const envp[]) +{ + /* Avoid the HAE being gratuitously wrong, which would cause us + to do the whole turn off interrupts thing and restore it. */ + struct pt_regs regs = {.hae = alpha_mv.hae_cache}; + int err = do_execve(path, argv, envp, ®s); + if (!err) { + struct pt_regs *p = current_pt_regs(); + /* copy regs to normal position and off to userland we go... */ + *p = regs; + __asm__ __volatile__ ( + "mov %0, $sp;" + "br $31, ret_from_sys_call" + : : "r"(p)); + } + return err; +} +EXPORT_SYMBOL(kernel_execve); diff --git a/arch/alpha/kernel/systbls.S b/arch/alpha/kernel/systbls.S index 87835235f114..2ac6b45c3e00 100644 --- a/arch/alpha/kernel/systbls.S +++ b/arch/alpha/kernel/systbls.S @@ -111,7 +111,7 @@ sys_call_table: .quad sys_socket .quad sys_connect .quad sys_accept - .quad osf_getpriority /* 100 */ + .quad sys_osf_getpriority /* 100 */ .quad sys_send .quad sys_recv .quad sys_sigreturn @@ -522,6 +522,8 @@ sys_call_table: .quad sys_setns .quad sys_accept4 .quad sys_sendmmsg + .quad sys_process_vm_readv + .quad sys_process_vm_writev /* 505 */ .size sys_call_table, . - sys_call_table .type sys_call_table, @object diff --git a/arch/alpha/lib/Makefile b/arch/alpha/lib/Makefile index c0a83ab62b78..59660743237c 100644 --- a/arch/alpha/lib/Makefile +++ b/arch/alpha/lib/Makefile @@ -31,8 +31,6 @@ lib-y = __divqu.o __remqu.o __divlu.o __remlu.o \ $(ev6-y)memchr.o \ $(ev6-y)copy_user.o \ $(ev6-y)clear_user.o \ - $(ev6-y)strncpy_from_user.o \ - $(ev67-y)strlen_user.o \ $(ev6-y)csum_ipv6_magic.o \ $(ev6-y)clear_page.o \ $(ev6-y)copy_page.o \ diff --git a/arch/alpha/lib/ev6-strncpy_from_user.S b/arch/alpha/lib/ev6-strncpy_from_user.S deleted file mode 100644 index d2e28178cacc..000000000000 --- a/arch/alpha/lib/ev6-strncpy_from_user.S +++ /dev/null @@ -1,424 +0,0 @@ -/* - * arch/alpha/lib/ev6-strncpy_from_user.S - * 21264 version contributed by Rick Gorton <rick.gorton@alpha-processor.com> - * - * Just like strncpy except in the return value: - * - * -EFAULT if an exception occurs before the terminator is copied. - * N if the buffer filled. - * - * Otherwise the length of the string is returned. - * - * Much of the information about 21264 scheduling/coding comes from: - * Compiler Writer's Guide for the Alpha 21264 - * abbreviated as 'CWG' in other comments here - * ftp.digital.com/pub/Digital/info/semiconductor/literature/dsc-library.html - * Scheduling notation: - * E - either cluster - * U - upper subcluster; U0 - subcluster U0; U1 - subcluster U1 - * L - lower subcluster; L0 - subcluster L0; L1 - subcluster L1 - * A bunch of instructions got moved and temp registers were changed - * to aid in scheduling. Control flow was also re-arranged to eliminate - * branches, and to provide longer code sequences to enable better scheduling. - * A total rewrite (using byte load/stores for start & tail sequences) - * is desirable, but very difficult to do without a from-scratch rewrite. - * Save that for the future. - */ - - -#include <asm/errno.h> -#include <asm/regdef.h> - - -/* Allow an exception for an insn; exit if we get one. */ -#define EX(x,y...) \ - 99: x,##y; \ - .section __ex_table,"a"; \ - .long 99b - .; \ - lda $31, $exception-99b($0); \ - .previous - - - .set noat - .set noreorder - .text - - .globl __strncpy_from_user - .ent __strncpy_from_user - .frame $30, 0, $26 - .prologue 0 - - .align 4 -__strncpy_from_user: - and a0, 7, t3 # E : find dest misalignment - beq a2, $zerolength # U : - - /* Are source and destination co-aligned? */ - mov a0, v0 # E : save the string start - xor a0, a1, t4 # E : - EX( ldq_u t1, 0(a1) ) # L : Latency=3 load first quadword - ldq_u t0, 0(a0) # L : load first (partial) aligned dest quadword - - addq a2, t3, a2 # E : bias count by dest misalignment - subq a2, 1, a3 # E : - addq zero, 1, t10 # E : - and t4, 7, t4 # E : misalignment between the two - - and a3, 7, t6 # E : number of tail bytes - sll t10, t6, t10 # E : t10 = bitmask of last count byte - bne t4, $unaligned # U : - lda t2, -1 # E : build a mask against false zero - - /* - * We are co-aligned; take care of a partial first word. - * On entry to this basic block: - * t0 == the first destination word for masking back in - * t1 == the first source word. - */ - - srl a3, 3, a2 # E : a2 = loop counter = (count - 1)/8 - addq a1, 8, a1 # E : - mskqh t2, a1, t2 # U : detection in the src word - nop - - /* Create the 1st output word and detect 0's in the 1st input word. */ - mskqh t1, a1, t3 # U : - mskql t0, a1, t0 # U : assemble the first output word - ornot t1, t2, t2 # E : - nop - - cmpbge zero, t2, t8 # E : bits set iff null found - or t0, t3, t0 # E : - beq a2, $a_eoc # U : - bne t8, $a_eos # U : 2nd branch in a quad. Bad. - - /* On entry to this basic block: - * t0 == a source quad not containing a null. - * a0 - current aligned destination address - * a1 - current aligned source address - * a2 - count of quadwords to move. - * NOTE: Loop improvement - unrolling this is going to be - * a huge win, since we're going to stall otherwise. - * Fix this later. For _really_ large copies, look - * at using wh64 on a look-ahead basis. See the code - * in clear_user.S and copy_user.S. - * Presumably, since (a0) and (a1) do not overlap (by C definition) - * Lots of nops here: - * - Separate loads from stores - * - Keep it to 1 branch/quadpack so the branch predictor - * can train. - */ -$a_loop: - stq_u t0, 0(a0) # L : - addq a0, 8, a0 # E : - nop - subq a2, 1, a2 # E : - - EX( ldq_u t0, 0(a1) ) # L : - addq a1, 8, a1 # E : - cmpbge zero, t0, t8 # E : Stall 2 cycles on t0 - beq a2, $a_eoc # U : - - beq t8, $a_loop # U : - nop - nop - nop - - /* Take care of the final (partial) word store. At this point - * the end-of-count bit is set in t8 iff it applies. - * - * On entry to this basic block we have: - * t0 == the source word containing the null - * t8 == the cmpbge mask that found it. - */ -$a_eos: - negq t8, t12 # E : find low bit set - and t8, t12, t12 # E : - - /* We're doing a partial word store and so need to combine - our source and original destination words. */ - ldq_u t1, 0(a0) # L : - subq t12, 1, t6 # E : - - or t12, t6, t8 # E : - zapnot t0, t8, t0 # U : clear src bytes > null - zap t1, t8, t1 # U : clear dst bytes <= null - or t0, t1, t0 # E : - - stq_u t0, 0(a0) # L : - br $finish_up # L0 : - nop - nop - - /* Add the end-of-count bit to the eos detection bitmask. */ - .align 4 -$a_eoc: - or t10, t8, t8 - br $a_eos - nop - nop - - -/* The source and destination are not co-aligned. Align the destination - and cope. We have to be very careful about not reading too much and - causing a SEGV. */ - - .align 4 -$u_head: - /* We know just enough now to be able to assemble the first - full source word. We can still find a zero at the end of it - that prevents us from outputting the whole thing. - - On entry to this basic block: - t0 == the first dest word, unmasked - t1 == the shifted low bits of the first source word - t6 == bytemask that is -1 in dest word bytes */ - - EX( ldq_u t2, 8(a1) ) # L : load second src word - addq a1, 8, a1 # E : - mskql t0, a0, t0 # U : mask trailing garbage in dst - extqh t2, a1, t4 # U : - - or t1, t4, t1 # E : first aligned src word complete - mskqh t1, a0, t1 # U : mask leading garbage in src - or t0, t1, t0 # E : first output word complete - or t0, t6, t6 # E : mask original data for zero test - - cmpbge zero, t6, t8 # E : - beq a2, $u_eocfin # U : - bne t8, $u_final # U : bad news - 2nd branch in a quad - lda t6, -1 # E : mask out the bits we have - - mskql t6, a1, t6 # U : already seen - stq_u t0, 0(a0) # L : store first output word - or t6, t2, t2 # E : - cmpbge zero, t2, t8 # E : find nulls in second partial - - addq a0, 8, a0 # E : - subq a2, 1, a2 # E : - bne t8, $u_late_head_exit # U : - nop - - /* Finally, we've got all the stupid leading edge cases taken care - of and we can set up to enter the main loop. */ - - extql t2, a1, t1 # U : position hi-bits of lo word - EX( ldq_u t2, 8(a1) ) # L : read next high-order source word - addq a1, 8, a1 # E : - cmpbge zero, t2, t8 # E : - - beq a2, $u_eoc # U : - bne t8, $u_eos # U : - nop - nop - - /* Unaligned copy main loop. In order to avoid reading too much, - the loop is structured to detect zeros in aligned source words. - This has, unfortunately, effectively pulled half of a loop - iteration out into the head and half into the tail, but it does - prevent nastiness from accumulating in the very thing we want - to run as fast as possible. - - On entry to this basic block: - t1 == the shifted high-order bits from the previous source word - t2 == the unshifted current source word - - We further know that t2 does not contain a null terminator. */ - - /* - * Extra nops here: - * separate load quads from store quads - * only one branch/quad to permit predictor training - */ - - .align 4 -$u_loop: - extqh t2, a1, t0 # U : extract high bits for current word - addq a1, 8, a1 # E : - extql t2, a1, t3 # U : extract low bits for next time - addq a0, 8, a0 # E : - - or t0, t1, t0 # E : current dst word now complete - EX( ldq_u t2, 0(a1) ) # L : load high word for next time - subq a2, 1, a2 # E : - nop - - stq_u t0, -8(a0) # L : save the current word - mov t3, t1 # E : - cmpbge zero, t2, t8 # E : test new word for eos - beq a2, $u_eoc # U : - - beq t8, $u_loop # U : - nop - nop - nop - - /* We've found a zero somewhere in the source word we just read. - If it resides in the lower half, we have one (probably partial) - word to write out, and if it resides in the upper half, we - have one full and one partial word left to write out. - - On entry to this basic block: - t1 == the shifted high-order bits from the previous source word - t2 == the unshifted current source word. */ - .align 4 -$u_eos: - extqh t2, a1, t0 # U : - or t0, t1, t0 # E : first (partial) source word complete - cmpbge zero, t0, t8 # E : is the null in this first bit? - nop - - bne t8, $u_final # U : - stq_u t0, 0(a0) # L : the null was in the high-order bits - addq a0, 8, a0 # E : - subq a2, 1, a2 # E : - - .align 4 -$u_late_head_exit: - extql t2, a1, t0 # U : - cmpbge zero, t0, t8 # E : - or t8, t10, t6 # E : - cmoveq a2, t6, t8 # E : - - /* Take care of a final (probably partial) result word. - On entry to this basic block: - t0 == assembled source word - t8 == cmpbge mask that found the null. */ - .align 4 -$u_final: - negq t8, t6 # E : isolate low bit set - and t6, t8, t12 # E : - ldq_u t1, 0(a0) # L : - subq t12, 1, t6 # E : - - or t6, t12, t8 # E : - zapnot t0, t8, t0 # U : kill source bytes > null - zap t1, t8, t1 # U : kill dest bytes <= null - or t0, t1, t0 # E : - - stq_u t0, 0(a0) # E : - br $finish_up # U : - nop - nop - - .align 4 -$u_eoc: # end-of-count - extqh t2, a1, t0 # U : - or t0, t1, t0 # E : - cmpbge zero, t0, t8 # E : - nop - - .align 4 -$u_eocfin: # end-of-count, final word - or t10, t8, t8 # E : - br $u_final # U : - nop - nop - - /* Unaligned copy entry point. */ - .align 4 -$unaligned: - - srl a3, 3, a2 # U : a2 = loop counter = (count - 1)/8 - and a0, 7, t4 # E : find dest misalignment - and a1, 7, t5 # E : find src misalignment - mov zero, t0 # E : - - /* Conditionally load the first destination word and a bytemask - with 0xff indicating that the destination byte is sacrosanct. */ - - mov zero, t6 # E : - beq t4, 1f # U : - ldq_u t0, 0(a0) # L : - lda t6, -1 # E : - - mskql t6, a0, t6 # E : - nop - nop - nop - - .align 4 -1: - subq a1, t4, a1 # E : sub dest misalignment from src addr - /* If source misalignment is larger than dest misalignment, we need - extra startup checks to avoid SEGV. */ - cmplt t4, t5, t12 # E : - extql t1, a1, t1 # U : shift src into place - lda t2, -1 # E : for creating masks later - - beq t12, $u_head # U : - mskqh t2, t5, t2 # U : begin src byte validity mask - cmpbge zero, t1, t8 # E : is there a zero? - nop - - extql t2, a1, t2 # U : - or t8, t10, t5 # E : test for end-of-count too - cmpbge zero, t2, t3 # E : - cmoveq a2, t5, t8 # E : Latency=2, extra map slot - - nop # E : goes with cmov - andnot t8, t3, t8 # E : - beq t8, $u_head # U : - nop - - /* At this point we've found a zero in the first partial word of - the source. We need to isolate the valid source data and mask - it into the original destination data. (Incidentally, we know - that we'll need at least one byte of that original dest word.) */ - - ldq_u t0, 0(a0) # L : - negq t8, t6 # E : build bitmask of bytes <= zero - mskqh t1, t4, t1 # U : - and t6, t8, t12 # E : - - subq t12, 1, t6 # E : - or t6, t12, t8 # E : - zapnot t2, t8, t2 # U : prepare source word; mirror changes - zapnot t1, t8, t1 # U : to source validity mask - - andnot t0, t2, t0 # E : zero place for source to reside - or t0, t1, t0 # E : and put it there - stq_u t0, 0(a0) # L : - nop - - .align 4 -$finish_up: - zapnot t0, t12, t4 # U : was last byte written null? - and t12, 0xf0, t3 # E : binary search for the address of the - cmovne t4, 1, t4 # E : Latency=2, extra map slot - nop # E : with cmovne - - and t12, 0xcc, t2 # E : last byte written - and t12, 0xaa, t1 # E : - cmovne t3, 4, t3 # E : Latency=2, extra map slot - nop # E : with cmovne - - bic a0, 7, t0 - cmovne t2, 2, t2 # E : Latency=2, extra map slot - nop # E : with cmovne - nop - - cmovne t1, 1, t1 # E : Latency=2, extra map slot - nop # E : with cmovne - addq t0, t3, t0 # E : - addq t1, t2, t1 # E : - - addq t0, t1, t0 # E : - addq t0, t4, t0 # add one if we filled the buffer - subq t0, v0, v0 # find string length - ret # L0 : - - .align 4 -$zerolength: - nop - nop - nop - clr v0 - -$exception: - nop - nop - nop - ret - - .end __strncpy_from_user diff --git a/arch/alpha/lib/ev67-strlen_user.S b/arch/alpha/lib/ev67-strlen_user.S deleted file mode 100644 index 57e0d77b81a6..000000000000 --- a/arch/alpha/lib/ev67-strlen_user.S +++ /dev/null @@ -1,107 +0,0 @@ -/* - * arch/alpha/lib/ev67-strlen_user.S - * 21264 version contributed by Rick Gorton <rick.gorton@api-networks.com> - * - * Return the length of the string including the NULL terminator - * (strlen+1) or zero if an error occurred. - * - * In places where it is critical to limit the processing time, - * and the data is not trusted, strnlen_user() should be used. - * It will return a value greater than its second argument if - * that limit would be exceeded. This implementation is allowed - * to access memory beyond the limit, but will not cross a page - * boundary when doing so. - * - * Much of the information about 21264 scheduling/coding comes from: - * Compiler Writer's Guide for the Alpha 21264 - * abbreviated as 'CWG' in other comments here - * ftp.digital.com/pub/Digital/info/semiconductor/literature/dsc-library.html - * Scheduling notation: - * E - either cluster - * U - upper subcluster; U0 - subcluster U0; U1 - subcluster U1 - * L - lower subcluster; L0 - subcluster L0; L1 - subcluster L1 - * Try not to change the actual algorithm if possible for consistency. - */ - -#include <asm/regdef.h> - - -/* Allow an exception for an insn; exit if we get one. */ -#define EX(x,y...) \ - 99: x,##y; \ - .section __ex_table,"a"; \ - .long 99b - .; \ - lda v0, $exception-99b(zero); \ - .previous - - - .set noreorder - .set noat - .text - - .globl __strlen_user - .ent __strlen_user - .frame sp, 0, ra - - .align 4 -__strlen_user: - ldah a1, 32767(zero) # do not use plain strlen_user() for strings - # that might be almost 2 GB long; you should - # be using strnlen_user() instead - nop - nop - nop - - .globl __strnlen_user - - .align 4 -__strnlen_user: - .prologue 0 - EX( ldq_u t0, 0(a0) ) # L : load first quadword (a0 may be misaligned) - lda t1, -1(zero) # E : - - insqh t1, a0, t1 # U : - andnot a0, 7, v0 # E : - or t1, t0, t0 # E : - subq a0, 1, a0 # E : get our +1 for the return - - cmpbge zero, t0, t1 # E : t1 <- bitmask: bit i == 1 <==> i-th byte == 0 - subq a1, 7, t2 # E : - subq a0, v0, t0 # E : - bne t1, $found # U : - - addq t2, t0, t2 # E : - addq a1, 1, a1 # E : - nop # E : - nop # E : - - .align 4 -$loop: ble t2, $limit # U : - EX( ldq t0, 8(v0) ) # L : - nop # E : - nop # E : - - cmpbge zero, t0, t1 # E : - subq t2, 8, t2 # E : - addq v0, 8, v0 # E : addr += 8 - beq t1, $loop # U : - -$found: cttz t1, t2 # U0 : - addq v0, t2, v0 # E : - subq v0, a0, v0 # E : - ret # L0 : - -$exception: - nop - nop - nop - ret - - .align 4 # currently redundant -$limit: - nop - nop - subq a1, t2, v0 - ret - - .end __strlen_user diff --git a/arch/alpha/lib/strlen_user.S b/arch/alpha/lib/strlen_user.S deleted file mode 100644 index 508a18e96479..000000000000 --- a/arch/alpha/lib/strlen_user.S +++ /dev/null @@ -1,91 +0,0 @@ -/* - * arch/alpha/lib/strlen_user.S - * - * Return the length of the string including the NUL terminator - * (strlen+1) or zero if an error occurred. - * - * In places where it is critical to limit the processing time, - * and the data is not trusted, strnlen_user() should be used. - * It will return a value greater than its second argument if - * that limit would be exceeded. This implementation is allowed - * to access memory beyond the limit, but will not cross a page - * boundary when doing so. - */ - -#include <asm/regdef.h> - - -/* Allow an exception for an insn; exit if we get one. */ -#define EX(x,y...) \ - 99: x,##y; \ - .section __ex_table,"a"; \ - .long 99b - .; \ - lda v0, $exception-99b(zero); \ - .previous - - - .set noreorder - .set noat - .text - - .globl __strlen_user - .ent __strlen_user - .frame sp, 0, ra - - .align 3 -__strlen_user: - ldah a1, 32767(zero) # do not use plain strlen_user() for strings - # that might be almost 2 GB long; you should - # be using strnlen_user() instead - - .globl __strnlen_user - - .align 3 -__strnlen_user: - .prologue 0 - - EX( ldq_u t0, 0(a0) ) # load first quadword (a0 may be misaligned) - lda t1, -1(zero) - insqh t1, a0, t1 - andnot a0, 7, v0 - or t1, t0, t0 - subq a0, 1, a0 # get our +1 for the return - cmpbge zero, t0, t1 # t1 <- bitmask: bit i == 1 <==> i-th byte == 0 - subq a1, 7, t2 - subq a0, v0, t0 - bne t1, $found - - addq t2, t0, t2 - addq a1, 1, a1 - - .align 3 -$loop: ble t2, $limit - EX( ldq t0, 8(v0) ) - subq t2, 8, t2 - addq v0, 8, v0 # addr += 8 - cmpbge zero, t0, t1 - beq t1, $loop - -$found: negq t1, t2 # clear all but least set bit - and t1, t2, t1 - - and t1, 0xf0, t2 # binary search for that set bit - and t1, 0xcc, t3 - and t1, 0xaa, t4 - cmovne t2, 4, t2 - cmovne t3, 2, t3 - cmovne t4, 1, t4 - addq t2, t3, t2 - addq v0, t4, v0 - addq v0, t2, v0 - nop # dual issue next two on ev4 and ev5 - subq v0, a0, v0 -$exception: - ret - - .align 3 # currently redundant -$limit: - subq a1, t2, v0 - ret - - .end __strlen_user diff --git a/arch/alpha/lib/strncpy_from_user.S b/arch/alpha/lib/strncpy_from_user.S deleted file mode 100644 index 73ee21160ff7..000000000000 --- a/arch/alpha/lib/strncpy_from_user.S +++ /dev/null @@ -1,339 +0,0 @@ -/* - * arch/alpha/lib/strncpy_from_user.S - * Contributed by Richard Henderson (rth@tamu.edu) - * - * Just like strncpy except in the return value: - * - * -EFAULT if an exception occurs before the terminator is copied. - * N if the buffer filled. - * - * Otherwise the length of the string is returned. - */ - - -#include <asm/errno.h> -#include <asm/regdef.h> - - -/* Allow an exception for an insn; exit if we get one. */ -#define EX(x,y...) \ - 99: x,##y; \ - .section __ex_table,"a"; \ - .long 99b - .; \ - lda $31, $exception-99b($0); \ - .previous - - - .set noat - .set noreorder - .text - - .globl __strncpy_from_user - .ent __strncpy_from_user - .frame $30, 0, $26 - .prologue 0 - - .align 3 -$aligned: - /* On entry to this basic block: - t0 == the first destination word for masking back in - t1 == the first source word. */ - - /* Create the 1st output word and detect 0's in the 1st input word. */ - lda t2, -1 # e1 : build a mask against false zero - mskqh t2, a1, t2 # e0 : detection in the src word - mskqh t1, a1, t3 # e0 : - ornot t1, t2, t2 # .. e1 : - mskql t0, a1, t0 # e0 : assemble the first output word - cmpbge zero, t2, t8 # .. e1 : bits set iff null found - or t0, t3, t0 # e0 : - beq a2, $a_eoc # .. e1 : - bne t8, $a_eos # .. e1 : - - /* On entry to this basic block: - t0 == a source word not containing a null. */ - -$a_loop: - stq_u t0, 0(a0) # e0 : - addq a0, 8, a0 # .. e1 : - EX( ldq_u t0, 0(a1) ) # e0 : - addq a1, 8, a1 # .. e1 : - subq a2, 1, a2 # e0 : - cmpbge zero, t0, t8 # .. e1 (stall) - beq a2, $a_eoc # e1 : - beq t8, $a_loop # e1 : - - /* Take care of the final (partial) word store. At this point - the end-of-count bit is set in t8 iff it applies. - - On entry to this basic block we have: - t0 == the source word containing the null - t8 == the cmpbge mask that found it. */ - -$a_eos: - negq t8, t12 # e0 : find low bit set - and t8, t12, t12 # e1 (stall) - - /* For the sake of the cache, don't read a destination word - if we're not going to need it. */ - and t12, 0x80, t6 # e0 : - bne t6, 1f # .. e1 (zdb) - - /* We're doing a partial word store and so need to combine - our source and original destination words. */ - ldq_u t1, 0(a0) # e0 : - subq t12, 1, t6 # .. e1 : - or t12, t6, t8 # e0 : - unop # - zapnot t0, t8, t0 # e0 : clear src bytes > null - zap t1, t8, t1 # .. e1 : clear dst bytes <= null - or t0, t1, t0 # e1 : - -1: stq_u t0, 0(a0) - br $finish_up - - /* Add the end-of-count bit to the eos detection bitmask. */ -$a_eoc: - or t10, t8, t8 - br $a_eos - - /*** The Function Entry Point ***/ - .align 3 -__strncpy_from_user: - mov a0, v0 # save the string start - beq a2, $zerolength - - /* Are source and destination co-aligned? */ - xor a0, a1, t1 # e0 : - and a0, 7, t0 # .. e1 : find dest misalignment - and t1, 7, t1 # e0 : - addq a2, t0, a2 # .. e1 : bias count by dest misalignment - subq a2, 1, a2 # e0 : - and a2, 7, t2 # e1 : - srl a2, 3, a2 # e0 : a2 = loop counter = (count - 1)/8 - addq zero, 1, t10 # .. e1 : - sll t10, t2, t10 # e0 : t10 = bitmask of last count byte - bne t1, $unaligned # .. e1 : - - /* We are co-aligned; take care of a partial first word. */ - - EX( ldq_u t1, 0(a1) ) # e0 : load first src word - addq a1, 8, a1 # .. e1 : - - beq t0, $aligned # avoid loading dest word if not needed - ldq_u t0, 0(a0) # e0 : - br $aligned # .. e1 : - - -/* The source and destination are not co-aligned. Align the destination - and cope. We have to be very careful about not reading too much and - causing a SEGV. */ - - .align 3 -$u_head: - /* We know just enough now to be able to assemble the first - full source word. We can still find a zero at the end of it - that prevents us from outputting the whole thing. - - On entry to this basic block: - t0 == the first dest word, unmasked - t1 == the shifted low bits of the first source word - t6 == bytemask that is -1 in dest word bytes */ - - EX( ldq_u t2, 8(a1) ) # e0 : load second src word - addq a1, 8, a1 # .. e1 : - mskql t0, a0, t0 # e0 : mask trailing garbage in dst - extqh t2, a1, t4 # e0 : - or t1, t4, t1 # e1 : first aligned src word complete - mskqh t1, a0, t1 # e0 : mask leading garbage in src - or t0, t1, t0 # e0 : first output word complete - or t0, t6, t6 # e1 : mask original data for zero test - cmpbge zero, t6, t8 # e0 : - beq a2, $u_eocfin # .. e1 : - bne t8, $u_final # e1 : - - lda t6, -1 # e1 : mask out the bits we have - mskql t6, a1, t6 # e0 : already seen - stq_u t0, 0(a0) # e0 : store first output word - or t6, t2, t2 # .. e1 : - cmpbge zero, t2, t8 # e0 : find nulls in second partial - addq a0, 8, a0 # .. e1 : - subq a2, 1, a2 # e0 : - bne t8, $u_late_head_exit # .. e1 : - - /* Finally, we've got all the stupid leading edge cases taken care - of and we can set up to enter the main loop. */ - - extql t2, a1, t1 # e0 : position hi-bits of lo word - EX( ldq_u t2, 8(a1) ) # .. e1 : read next high-order source word - addq a1, 8, a1 # e0 : - cmpbge zero, t2, t8 # e1 (stall) - beq a2, $u_eoc # e1 : - bne t8, $u_eos # e1 : - - /* Unaligned copy main loop. In order to avoid reading too much, - the loop is structured to detect zeros in aligned source words. - This has, unfortunately, effectively pulled half of a loop - iteration out into the head and half into the tail, but it does - prevent nastiness from accumulating in the very thing we want - to run as fast as possible. - - On entry to this basic block: - t1 == the shifted high-order bits from the previous source word - t2 == the unshifted current source word - - We further know that t2 does not contain a null terminator. */ - - .align 3 -$u_loop: - extqh t2, a1, t0 # e0 : extract high bits for current word - addq a1, 8, a1 # .. e1 : - extql t2, a1, t3 # e0 : extract low bits for next time - addq a0, 8, a0 # .. e1 : - or t0, t1, t0 # e0 : current dst word now complete - EX( ldq_u t2, 0(a1) ) # .. e1 : load high word for next time - stq_u t0, -8(a0) # e0 : save the current word - mov t3, t1 # .. e1 : - subq a2, 1, a2 # e0 : - cmpbge zero, t2, t8 # .. e1 : test new word for eos - beq a2, $u_eoc # e1 : - beq t8, $u_loop # e1 : - - /* We've found a zero somewhere in the source word we just read. - If it resides in the lower half, we have one (probably partial) - word to write out, and if it resides in the upper half, we - have one full and one partial word left to write out. - - On entry to this basic block: - t1 == the shifted high-order bits from the previous source word - t2 == the unshifted current source word. */ -$u_eos: - extqh t2, a1, t0 # e0 : - or t0, t1, t0 # e1 : first (partial) source word complete - - cmpbge zero, t0, t8 # e0 : is the null in this first bit? - bne t8, $u_final # .. e1 (zdb) - - stq_u t0, 0(a0) # e0 : the null was in the high-order bits - addq a0, 8, a0 # .. e1 : - subq a2, 1, a2 # e1 : - -$u_late_head_exit: - extql t2, a1, t0 # .. e0 : - cmpbge zero, t0, t8 # e0 : - or t8, t10, t6 # e1 : - cmoveq a2, t6, t8 # e0 : - nop # .. e1 : - - /* Take care of a final (probably partial) result word. - On entry to this basic block: - t0 == assembled source word - t8 == cmpbge mask that found the null. */ -$u_final: - negq t8, t6 # e0 : isolate low bit set - and t6, t8, t12 # e1 : - - and t12, 0x80, t6 # e0 : avoid dest word load if we can - bne t6, 1f # .. e1 (zdb) - - ldq_u t1, 0(a0) # e0 : - subq t12, 1, t6 # .. e1 : - or t6, t12, t8 # e0 : - zapnot t0, t8, t0 # .. e1 : kill source bytes > null - zap t1, t8, t1 # e0 : kill dest bytes <= null - or t0, t1, t0 # e1 : - -1: stq_u t0, 0(a0) # e0 : - br $finish_up - -$u_eoc: # end-of-count - extqh t2, a1, t0 - or t0, t1, t0 - cmpbge zero, t0, t8 - -$u_eocfin: # end-of-count, final word - or t10, t8, t8 - br $u_final - - /* Unaligned copy entry point. */ - .align 3 -$unaligned: - - EX( ldq_u t1, 0(a1) ) # e0 : load first source word - - and a0, 7, t4 # .. e1 : find dest misalignment - and a1, 7, t5 # e0 : find src misalignment - - /* Conditionally load the first destination word and a bytemask - with 0xff indicating that the destination byte is sacrosanct. */ - - mov zero, t0 # .. e1 : - mov zero, t6 # e0 : - beq t4, 1f # .. e1 : - ldq_u t0, 0(a0) # e0 : - lda t6, -1 # .. e1 : - mskql t6, a0, t6 # e0 : -1: - subq a1, t4, a1 # .. e1 : sub dest misalignment from src addr - - /* If source misalignment is larger than dest misalignment, we need - extra startup checks to avoid SEGV. */ - - cmplt t4, t5, t12 # e1 : - extql t1, a1, t1 # .. e0 : shift src into place - lda t2, -1 # e0 : for creating masks later - beq t12, $u_head # e1 : - - mskqh t2, t5, t2 # e0 : begin src byte validity mask - cmpbge zero, t1, t8 # .. e1 : is there a zero? - extql t2, a1, t2 # e0 : - or t8, t10, t5 # .. e1 : test for end-of-count too - cmpbge zero, t2, t3 # e0 : - cmoveq a2, t5, t8 # .. e1 : - andnot t8, t3, t8 # e0 : - beq t8, $u_head # .. e1 (zdb) - - /* At this point we've found a zero in the first partial word of - the source. We need to isolate the valid source data and mask - it into the original destination data. (Incidentally, we know - that we'll need at least one byte of that original dest word.) */ - - ldq_u t0, 0(a0) # e0 : - negq t8, t6 # .. e1 : build bitmask of bytes <= zero - mskqh t1, t4, t1 # e0 : - and t6, t8, t12 # .. e1 : - subq t12, 1, t6 # e0 : - or t6, t12, t8 # e1 : - - zapnot t2, t8, t2 # e0 : prepare source word; mirror changes - zapnot t1, t8, t1 # .. e1 : to source validity mask - - andnot t0, t2, t0 # e0 : zero place for source to reside - or t0, t1, t0 # e1 : and put it there - stq_u t0, 0(a0) # e0 : - -$finish_up: - zapnot t0, t12, t4 # was last byte written null? - cmovne t4, 1, t4 - - and t12, 0xf0, t3 # binary search for the address of the - and t12, 0xcc, t2 # last byte written - and t12, 0xaa, t1 - bic a0, 7, t0 - cmovne t3, 4, t3 - cmovne t2, 2, t2 - cmovne t1, 1, t1 - addq t0, t3, t0 - addq t1, t2, t1 - addq t0, t1, t0 - addq t0, t4, t0 # add one if we filled the buffer - - subq t0, v0, v0 # find string length - ret - -$zerolength: - clr v0 -$exception: - ret - - .end __strncpy_from_user diff --git a/arch/alpha/mm/fault.c b/arch/alpha/mm/fault.c index 5eecab1a84ef..0c4132dd3507 100644 --- a/arch/alpha/mm/fault.c +++ b/arch/alpha/mm/fault.c @@ -89,6 +89,8 @@ do_page_fault(unsigned long address, unsigned long mmcsr, const struct exception_table_entry *fixup; int fault, si_code = SEGV_MAPERR; siginfo_t info; + unsigned int flags = (FAULT_FLAG_ALLOW_RETRY | FAULT_FLAG_KILLABLE | + (cause > 0 ? FAULT_FLAG_WRITE : 0)); /* As of EV6, a load into $31/$f31 is a prefetch, and never faults (or is suppressed by the PALcode). Support that for older CPUs @@ -114,6 +116,7 @@ do_page_fault(unsigned long address, unsigned long mmcsr, goto vmalloc_fault; #endif +retry: down_read(&mm->mmap_sem); vma = find_vma(mm, address); if (!vma) @@ -144,8 +147,11 @@ do_page_fault(unsigned long address, unsigned long mmcsr, /* If for any reason at all we couldn't handle the fault, make sure we exit gracefully rather than endlessly redo the fault. */ - fault = handle_mm_fault(mm, vma, address, cause > 0 ? FAULT_FLAG_WRITE : 0); - up_read(&mm->mmap_sem); + fault = handle_mm_fault(mm, vma, address, flags); + + if ((fault & VM_FAULT_RETRY) && fatal_signal_pending(current)) + return; + if (unlikely(fault & VM_FAULT_ERROR)) { if (fault & VM_FAULT_OOM) goto out_of_memory; @@ -153,10 +159,26 @@ do_page_fault(unsigned long address, unsigned long mmcsr, goto do_sigbus; BUG(); } - if (fault & VM_FAULT_MAJOR) - current->maj_flt++; - else - current->min_flt++; + + if (flags & FAULT_FLAG_ALLOW_RETRY) { + if (fault & VM_FAULT_MAJOR) + current->maj_flt++; + else + current->min_flt++; + if (fault & VM_FAULT_RETRY) { + flags &= ~FAULT_FLAG_ALLOW_RETRY; + + /* No need to up_read(&mm->mmap_sem) as we would + * have already released it in __lock_page_or_retry + * in mm/filemap.c. + */ + + goto retry; + } + } + + up_read(&mm->mmap_sem); + return; /* Something tried to access memory that isn't in our memory map. @@ -186,12 +208,14 @@ do_page_fault(unsigned long address, unsigned long mmcsr, /* We ran out of memory, or some other thing happened to us that made us unable to handle the page fault gracefully. */ out_of_memory: + up_read(&mm->mmap_sem); if (!user_mode(regs)) goto no_context; pagefault_out_of_memory(); return; do_sigbus: + up_read(&mm->mmap_sem); /* Send a sigbus, regardless of whether we were in kernel or user mode. */ info.si_signo = SIGBUS; diff --git a/arch/alpha/oprofile/common.c b/arch/alpha/oprofile/common.c index a0a5d27aa215..b8ce18f485d3 100644 --- a/arch/alpha/oprofile/common.c +++ b/arch/alpha/oprofile/common.c @@ -12,6 +12,7 @@ #include <linux/smp.h> #include <linux/errno.h> #include <asm/ptrace.h> +#include <asm/special_insns.h> #include "op_impl.h" diff --git a/arch/ia64/configs/generic_defconfig b/arch/ia64/configs/generic_defconfig index 954d81e2e837..7913695b2fcb 100644 --- a/arch/ia64/configs/generic_defconfig +++ b/arch/ia64/configs/generic_defconfig @@ -234,5 +234,4 @@ CONFIG_CRYPTO_PCBC=m CONFIG_CRYPTO_MD5=y # CONFIG_CRYPTO_ANSI_CPRNG is not set CONFIG_CRC_T10DIF=y -CONFIG_MISC_DEVICES=y CONFIG_INTEL_IOMMU=y diff --git a/arch/ia64/configs/gensparse_defconfig b/arch/ia64/configs/gensparse_defconfig index 91c41ecfa6d9..f8e913365423 100644 --- a/arch/ia64/configs/gensparse_defconfig +++ b/arch/ia64/configs/gensparse_defconfig @@ -209,4 +209,3 @@ CONFIG_MAGIC_SYSRQ=y CONFIG_DEBUG_KERNEL=y CONFIG_DEBUG_MUTEXES=y CONFIG_CRYPTO_MD5=y -CONFIG_MISC_DEVICES=y diff --git a/arch/m68k/Kconfig b/arch/m68k/Kconfig index 4a469907f04a..b22df9410dce 100644 --- a/arch/m68k/Kconfig +++ b/arch/m68k/Kconfig @@ -5,6 +5,7 @@ config M68K select HAVE_AOUT if MMU select HAVE_GENERIC_HARDIRQS select GENERIC_IRQ_SHOW + select GENERIC_ATOMIC64 select ARCH_HAVE_NMI_SAFE_CMPXCHG if RMW_INSNS select GENERIC_CPU_DEVICES select GENERIC_STRNCPY_FROM_USER if MMU diff --git a/arch/m68k/Kconfig.cpu b/arch/m68k/Kconfig.cpu index 82068349a2bb..c4eb79edecec 100644 --- a/arch/m68k/Kconfig.cpu +++ b/arch/m68k/Kconfig.cpu @@ -28,6 +28,7 @@ config COLDFIRE select CPU_HAS_NO_BITFIELDS select CPU_HAS_NO_MULDIV64 select GENERIC_CSUM + select HAVE_CLK endchoice @@ -58,7 +59,6 @@ config MCPU32 config M68020 bool "68020 support" depends on MMU - select GENERIC_ATOMIC64 select CPU_HAS_ADDRESS_SPACES help If you anticipate running this kernel on a computer with a MC68020 @@ -69,7 +69,6 @@ config M68020 config M68030 bool "68030 support" depends on MMU && !MMU_SUN3 - select GENERIC_ATOMIC64 select CPU_HAS_ADDRESS_SPACES help If you anticipate running this kernel on a computer with a MC68030 @@ -79,7 +78,6 @@ config M68030 config M68040 bool "68040 support" depends on MMU && !MMU_SUN3 - select GENERIC_ATOMIC64 select CPU_HAS_ADDRESS_SPACES help If you anticipate running this kernel on a computer with a MC68LC040 @@ -90,7 +88,6 @@ config M68040 config M68060 bool "68060 support" depends on MMU && !MMU_SUN3 - select GENERIC_ATOMIC64 select CPU_HAS_ADDRESS_SPACES help If you anticipate running this kernel on a computer with a MC68060 diff --git a/arch/x86/Makefile b/arch/x86/Makefile index b0c5276861ec..682e9c210baa 100644 --- a/arch/x86/Makefile +++ b/arch/x86/Makefile @@ -27,6 +27,10 @@ ifeq ($(CONFIG_X86_32),y) KBUILD_CFLAGS += -msoft-float -mregparm=3 -freg-struct-return + # Never want PIC in a 32-bit kernel, prevent breakage with GCC built + # with nonstandard options + KBUILD_CFLAGS += -fno-pic + # prevent gcc from keeping the stack 16 byte aligned KBUILD_CFLAGS += $(call cc-option,-mpreferred-stack-boundary=2) diff --git a/arch/x86/boot/Makefile b/arch/x86/boot/Makefile index 5a747dd884db..f7535bedc33f 100644 --- a/arch/x86/boot/Makefile +++ b/arch/x86/boot/Makefile @@ -57,7 +57,7 @@ KBUILD_CFLAGS := $(LINUXINCLUDE) -g -Os -D_SETUP -D__KERNEL__ \ -Wall -Wstrict-prototypes \ -march=i386 -mregparm=3 \ -include $(srctree)/$(src)/code16gcc.h \ - -fno-strict-aliasing -fomit-frame-pointer \ + -fno-strict-aliasing -fomit-frame-pointer -fno-pic \ $(call cc-option, -ffreestanding) \ $(call cc-option, -fno-toplevel-reorder,\ $(call cc-option, -fno-unit-at-a-time)) \ diff --git a/arch/x86/kernel/apic/io_apic.c b/arch/x86/kernel/apic/io_apic.c index a6c64aaddf9a..c265593ec2cd 100644 --- a/arch/x86/kernel/apic/io_apic.c +++ b/arch/x86/kernel/apic/io_apic.c @@ -1356,6 +1356,16 @@ static void setup_ioapic_irq(unsigned int irq, struct irq_cfg *cfg, if (!IO_APIC_IRQ(irq)) return; + /* + * For legacy irqs, cfg->domain starts with cpu 0. Now that IO-APIC + * can handle this irq and the apic driver is finialized at this point, + * update the cfg->domain. + */ + if (irq < legacy_pic->nr_legacy_irqs && + cpumask_equal(cfg->domain, cpumask_of(0))) + apic->vector_allocation_domain(0, cfg->domain, + apic->target_cpus()); + if (assign_irq_vector(irq, cfg, apic->target_cpus())) return; diff --git a/arch/x86/kernel/cpu/common.c b/arch/x86/kernel/cpu/common.c index 46d8786d655e..a5fbc3c5fccc 100644 --- a/arch/x86/kernel/cpu/common.c +++ b/arch/x86/kernel/cpu/common.c @@ -144,6 +144,8 @@ static int __init x86_xsave_setup(char *s) { setup_clear_cpu_cap(X86_FEATURE_XSAVE); setup_clear_cpu_cap(X86_FEATURE_XSAVEOPT); + setup_clear_cpu_cap(X86_FEATURE_AVX); + setup_clear_cpu_cap(X86_FEATURE_AVX2); return 1; } __setup("noxsave", x86_xsave_setup); diff --git a/arch/x86/kernel/cpu/perf_event_intel.c b/arch/x86/kernel/cpu/perf_event_intel.c index 382366977d4c..7f2739e03e79 100644 --- a/arch/x86/kernel/cpu/perf_event_intel.c +++ b/arch/x86/kernel/cpu/perf_event_intel.c @@ -1522,8 +1522,16 @@ static struct perf_guest_switch_msr *intel_guest_get_msrs(int *nr) arr[0].msr = MSR_CORE_PERF_GLOBAL_CTRL; arr[0].host = x86_pmu.intel_ctrl & ~cpuc->intel_ctrl_guest_mask; arr[0].guest = x86_pmu.intel_ctrl & ~cpuc->intel_ctrl_host_mask; + /* + * If PMU counter has PEBS enabled it is not enough to disable counter + * on a guest entry since PEBS memory write can overshoot guest entry + * and corrupt guest memory. Disabling PEBS solves the problem. + */ + arr[1].msr = MSR_IA32_PEBS_ENABLE; + arr[1].host = cpuc->pebs_enabled; + arr[1].guest = 0; - *nr = 1; + *nr = 2; return arr; } diff --git a/arch/x86/kernel/cpu/perf_event_intel_uncore.c b/arch/x86/kernel/cpu/perf_event_intel_uncore.c index 7563fda9f033..0a5571080e74 100644 --- a/arch/x86/kernel/cpu/perf_event_intel_uncore.c +++ b/arch/x86/kernel/cpu/perf_event_intel_uncore.c @@ -796,7 +796,6 @@ static struct intel_uncore_type *nhm_msr_uncores[] = { DEFINE_UNCORE_FORMAT_ATTR(event5, event, "config:1-5"); DEFINE_UNCORE_FORMAT_ATTR(counter, counter, "config:6-7"); -DEFINE_UNCORE_FORMAT_ATTR(mm_cfg, mm_cfg, "config:63"); DEFINE_UNCORE_FORMAT_ATTR(match, match, "config1:0-63"); DEFINE_UNCORE_FORMAT_ATTR(mask, mask, "config2:0-63"); @@ -902,16 +901,21 @@ static struct attribute_group nhmex_uncore_cbox_format_group = { .attrs = nhmex_uncore_cbox_formats_attr, }; +/* msr offset for each instance of cbox */ +static unsigned nhmex_cbox_msr_offsets[] = { + 0x0, 0x80, 0x40, 0xc0, 0x20, 0xa0, 0x60, 0xe0, 0x240, 0x2c0, +}; + static struct intel_uncore_type nhmex_uncore_cbox = { .name = "cbox", .num_counters = 6, - .num_boxes = 8, + .num_boxes = 10, .perf_ctr_bits = 48, .event_ctl = NHMEX_C0_MSR_PMON_EV_SEL0, .perf_ctr = NHMEX_C0_MSR_PMON_CTR0, .event_mask = NHMEX_PMON_RAW_EVENT_MASK, .box_ctl = NHMEX_C0_MSR_PMON_GLOBAL_CTL, - .msr_offset = NHMEX_C_MSR_OFFSET, + .msr_offsets = nhmex_cbox_msr_offsets, .pair_ctr_ctl = 1, .ops = &nhmex_uncore_ops, .format_group = &nhmex_uncore_cbox_format_group @@ -1032,24 +1036,22 @@ static struct intel_uncore_type nhmex_uncore_bbox = { static int nhmex_sbox_hw_config(struct intel_uncore_box *box, struct perf_event *event) { - struct hw_perf_event_extra *reg1 = &event->hw.extra_reg; - struct hw_perf_event_extra *reg2 = &event->hw.branch_reg; + struct hw_perf_event *hwc = &event->hw; + struct hw_perf_event_extra *reg1 = &hwc->extra_reg; + struct hw_perf_event_extra *reg2 = &hwc->branch_reg; - if (event->attr.config & NHMEX_S_PMON_MM_CFG_EN) { - reg1->config = event->attr.config1; - reg2->config = event->attr.config2; - } else { - reg1->config = ~0ULL; - reg2->config = ~0ULL; - } + /* only TO_R_PROG_EV event uses the match/mask register */ + if ((hwc->config & NHMEX_PMON_CTL_EV_SEL_MASK) != + NHMEX_S_EVENT_TO_R_PROG_EV) + return 0; if (box->pmu->pmu_idx == 0) reg1->reg = NHMEX_S0_MSR_MM_CFG; else reg1->reg = NHMEX_S1_MSR_MM_CFG; - reg1->idx = 0; - + reg1->config = event->attr.config1; + reg2->config = event->attr.config2; return 0; } @@ -1059,8 +1061,8 @@ static void nhmex_sbox_msr_enable_event(struct intel_uncore_box *box, struct per struct hw_perf_event_extra *reg1 = &hwc->extra_reg; struct hw_perf_event_extra *reg2 = &hwc->branch_reg; - wrmsrl(reg1->reg, 0); - if (reg1->config != ~0ULL || reg2->config != ~0ULL) { + if (reg1->idx != EXTRA_REG_NONE) { + wrmsrl(reg1->reg, 0); wrmsrl(reg1->reg + 1, reg1->config); wrmsrl(reg1->reg + 2, reg2->config); wrmsrl(reg1->reg, NHMEX_S_PMON_MM_CFG_EN); @@ -1074,7 +1076,6 @@ static struct attribute *nhmex_uncore_sbox_formats_attr[] = { &format_attr_edge.attr, &format_attr_inv.attr, &format_attr_thresh8.attr, - &format_attr_mm_cfg.attr, &format_attr_match.attr, &format_attr_mask.attr, NULL, @@ -1142,6 +1143,9 @@ static struct extra_reg nhmex_uncore_mbox_extra_regs[] = { EVENT_EXTRA_END }; +/* Nehalem-EX or Westmere-EX ? */ +bool uncore_nhmex; + static bool nhmex_mbox_get_shared_reg(struct intel_uncore_box *box, int idx, u64 config) { struct intel_uncore_extra_reg *er; @@ -1171,18 +1175,29 @@ static bool nhmex_mbox_get_shared_reg(struct intel_uncore_box *box, int idx, u64 return false; /* mask of the shared fields */ - mask = NHMEX_M_PMON_ZDP_CTL_FVC_MASK; + if (uncore_nhmex) + mask = NHMEX_M_PMON_ZDP_CTL_FVC_MASK; + else + mask = WSMEX_M_PMON_ZDP_CTL_FVC_MASK; er = &box->shared_regs[EXTRA_REG_NHMEX_M_ZDP_CTL_FVC]; raw_spin_lock_irqsave(&er->lock, flags); /* add mask of the non-shared field if it's in use */ - if (__BITS_VALUE(atomic_read(&er->ref), idx, 8)) - mask |= NHMEX_M_PMON_ZDP_CTL_FVC_EVENT_MASK(idx); + if (__BITS_VALUE(atomic_read(&er->ref), idx, 8)) { + if (uncore_nhmex) + mask |= NHMEX_M_PMON_ZDP_CTL_FVC_EVENT_MASK(idx); + else + mask |= WSMEX_M_PMON_ZDP_CTL_FVC_EVENT_MASK(idx); + } if (!atomic_read(&er->ref) || !((er->config ^ config) & mask)) { atomic_add(1 << (idx * 8), &er->ref); - mask = NHMEX_M_PMON_ZDP_CTL_FVC_MASK | - NHMEX_M_PMON_ZDP_CTL_FVC_EVENT_MASK(idx); + if (uncore_nhmex) + mask = NHMEX_M_PMON_ZDP_CTL_FVC_MASK | + NHMEX_M_PMON_ZDP_CTL_FVC_EVENT_MASK(idx); + else + mask = WSMEX_M_PMON_ZDP_CTL_FVC_MASK | + WSMEX_M_PMON_ZDP_CTL_FVC_EVENT_MASK(idx); er->config &= ~mask; er->config |= (config & mask); ret = true; @@ -1216,7 +1231,10 @@ u64 nhmex_mbox_alter_er(struct perf_event *event, int new_idx, bool modify) /* get the non-shared control bits and shift them */ idx = orig_idx - EXTRA_REG_NHMEX_M_ZDP_CTL_FVC; - config &= NHMEX_M_PMON_ZDP_CTL_FVC_EVENT_MASK(idx); + if (uncore_nhmex) + config &= NHMEX_M_PMON_ZDP_CTL_FVC_EVENT_MASK(idx); + else + config &= WSMEX_M_PMON_ZDP_CTL_FVC_EVENT_MASK(idx); if (new_idx > orig_idx) { idx = new_idx - orig_idx; config <<= 3 * idx; @@ -1226,6 +1244,10 @@ u64 nhmex_mbox_alter_er(struct perf_event *event, int new_idx, bool modify) } /* add the shared control bits back */ + if (uncore_nhmex) + config |= NHMEX_M_PMON_ZDP_CTL_FVC_MASK & reg1->config; + else + config |= WSMEX_M_PMON_ZDP_CTL_FVC_MASK & reg1->config; config |= NHMEX_M_PMON_ZDP_CTL_FVC_MASK & reg1->config; if (modify) { /* adjust the main event selector */ @@ -1264,7 +1286,8 @@ again: } /* for the match/mask registers */ - if ((uncore_box_is_fake(box) || !reg2->alloc) && + if (reg2->idx != EXTRA_REG_NONE && + (uncore_box_is_fake(box) || !reg2->alloc) && !nhmex_mbox_get_shared_reg(box, reg2->idx, reg2->config)) goto fail; @@ -1278,7 +1301,8 @@ again: if (idx[0] != 0xff && idx[0] != __BITS_VALUE(reg1->idx, 0, 8)) nhmex_mbox_alter_er(event, idx[0], true); reg1->alloc |= alloc; - reg2->alloc = 1; + if (reg2->idx != EXTRA_REG_NONE) + reg2->alloc = 1; } return NULL; fail: @@ -1342,9 +1366,6 @@ static int nhmex_mbox_hw_config(struct intel_uncore_box *box, struct perf_event struct extra_reg *er; unsigned msr; int reg_idx = 0; - - if (WARN_ON_ONCE(reg1->idx != -1)) - return -EINVAL; /* * The mbox events may require 2 extra MSRs at the most. But only * the lower 32 bits in these MSRs are significant, so we can use @@ -1355,11 +1376,6 @@ static int nhmex_mbox_hw_config(struct intel_uncore_box *box, struct perf_event continue; if (event->attr.config1 & ~er->valid_mask) return -EINVAL; - if (er->idx == __BITS_VALUE(reg1->idx, 0, 8) || - er->idx == __BITS_VALUE(reg1->idx, 1, 8)) - continue; - if (WARN_ON_ONCE(reg_idx >= 2)) - return -EINVAL; msr = er->msr + type->msr_offset * box->pmu->pmu_idx; if (WARN_ON_ONCE(msr >= 0xffff || er->idx >= 0xff)) @@ -1368,6 +1384,8 @@ static int nhmex_mbox_hw_config(struct intel_uncore_box *box, struct perf_event /* always use the 32~63 bits to pass the PLD config */ if (er->idx == EXTRA_REG_NHMEX_M_PLD) reg_idx = 1; + else if (WARN_ON_ONCE(reg_idx > 0)) + return -EINVAL; reg1->idx &= ~(0xff << (reg_idx * 8)); reg1->reg &= ~(0xffff << (reg_idx * 16)); @@ -1376,17 +1394,21 @@ static int nhmex_mbox_hw_config(struct intel_uncore_box *box, struct perf_event reg1->config = event->attr.config1; reg_idx++; } - /* use config2 to pass the filter config */ - reg2->idx = EXTRA_REG_NHMEX_M_FILTER; - if (event->attr.config2 & NHMEX_M_PMON_MM_CFG_EN) - reg2->config = event->attr.config2; - else - reg2->config = ~0ULL; - if (box->pmu->pmu_idx == 0) - reg2->reg = NHMEX_M0_MSR_PMU_MM_CFG; - else - reg2->reg = NHMEX_M1_MSR_PMU_MM_CFG; - + /* + * The mbox only provides ability to perform address matching + * for the PLD events. + */ + if (reg_idx == 2) { + reg2->idx = EXTRA_REG_NHMEX_M_FILTER; + if (event->attr.config2 & NHMEX_M_PMON_MM_CFG_EN) + reg2->config = event->attr.config2; + else + reg2->config = ~0ULL; + if (box->pmu->pmu_idx == 0) + reg2->reg = NHMEX_M0_MSR_PMU_MM_CFG; + else + reg2->reg = NHMEX_M1_MSR_PMU_MM_CFG; + } return 0; } @@ -1422,34 +1444,36 @@ static void nhmex_mbox_msr_enable_event(struct intel_uncore_box *box, struct per wrmsrl(__BITS_VALUE(reg1->reg, 1, 16), nhmex_mbox_shared_reg_config(box, idx)); - wrmsrl(reg2->reg, 0); - if (reg2->config != ~0ULL) { - wrmsrl(reg2->reg + 1, - reg2->config & NHMEX_M_PMON_ADDR_MATCH_MASK); - wrmsrl(reg2->reg + 2, NHMEX_M_PMON_ADDR_MASK_MASK & - (reg2->config >> NHMEX_M_PMON_ADDR_MASK_SHIFT)); - wrmsrl(reg2->reg, NHMEX_M_PMON_MM_CFG_EN); + if (reg2->idx != EXTRA_REG_NONE) { + wrmsrl(reg2->reg, 0); + if (reg2->config != ~0ULL) { + wrmsrl(reg2->reg + 1, + reg2->config & NHMEX_M_PMON_ADDR_MATCH_MASK); + wrmsrl(reg2->reg + 2, NHMEX_M_PMON_ADDR_MASK_MASK & + (reg2->config >> NHMEX_M_PMON_ADDR_MASK_SHIFT)); + wrmsrl(reg2->reg, NHMEX_M_PMON_MM_CFG_EN); + } } wrmsrl(hwc->config_base, hwc->config | NHMEX_PMON_CTL_EN_BIT0); } -DEFINE_UNCORE_FORMAT_ATTR(count_mode, count_mode, "config:2-3"); -DEFINE_UNCORE_FORMAT_ATTR(storage_mode, storage_mode, "config:4-5"); -DEFINE_UNCORE_FORMAT_ATTR(wrap_mode, wrap_mode, "config:6"); -DEFINE_UNCORE_FORMAT_ATTR(flag_mode, flag_mode, "config:7"); -DEFINE_UNCORE_FORMAT_ATTR(inc_sel, inc_sel, "config:9-13"); -DEFINE_UNCORE_FORMAT_ATTR(set_flag_sel, set_flag_sel, "config:19-21"); -DEFINE_UNCORE_FORMAT_ATTR(filter_cfg, filter_cfg, "config2:63"); -DEFINE_UNCORE_FORMAT_ATTR(filter_match, filter_match, "config2:0-33"); -DEFINE_UNCORE_FORMAT_ATTR(filter_mask, filter_mask, "config2:34-61"); -DEFINE_UNCORE_FORMAT_ATTR(dsp, dsp, "config1:0-31"); -DEFINE_UNCORE_FORMAT_ATTR(thr, thr, "config1:0-31"); -DEFINE_UNCORE_FORMAT_ATTR(fvc, fvc, "config1:0-31"); -DEFINE_UNCORE_FORMAT_ATTR(pgt, pgt, "config1:0-31"); -DEFINE_UNCORE_FORMAT_ATTR(map, map, "config1:0-31"); -DEFINE_UNCORE_FORMAT_ATTR(iss, iss, "config1:0-31"); -DEFINE_UNCORE_FORMAT_ATTR(pld, pld, "config1:32-63"); +DEFINE_UNCORE_FORMAT_ATTR(count_mode, count_mode, "config:2-3"); +DEFINE_UNCORE_FORMAT_ATTR(storage_mode, storage_mode, "config:4-5"); +DEFINE_UNCORE_FORMAT_ATTR(wrap_mode, wrap_mode, "config:6"); +DEFINE_UNCORE_FORMAT_ATTR(flag_mode, flag_mode, "config:7"); +DEFINE_UNCORE_FORMAT_ATTR(inc_sel, inc_sel, "config:9-13"); +DEFINE_UNCORE_FORMAT_ATTR(set_flag_sel, set_flag_sel, "config:19-21"); +DEFINE_UNCORE_FORMAT_ATTR(filter_cfg_en, filter_cfg_en, "config2:63"); +DEFINE_UNCORE_FORMAT_ATTR(filter_match, filter_match, "config2:0-33"); +DEFINE_UNCORE_FORMAT_ATTR(filter_mask, filter_mask, "config2:34-61"); +DEFINE_UNCORE_FORMAT_ATTR(dsp, dsp, "config1:0-31"); +DEFINE_UNCORE_FORMAT_ATTR(thr, thr, "config1:0-31"); +DEFINE_UNCORE_FORMAT_ATTR(fvc, fvc, "config1:0-31"); +DEFINE_UNCORE_FORMAT_ATTR(pgt, pgt, "config1:0-31"); +DEFINE_UNCORE_FORMAT_ATTR(map, map, "config1:0-31"); +DEFINE_UNCORE_FORMAT_ATTR(iss, iss, "config1:0-31"); +DEFINE_UNCORE_FORMAT_ATTR(pld, pld, "config1:32-63"); static struct attribute *nhmex_uncore_mbox_formats_attr[] = { &format_attr_count_mode.attr, @@ -1458,7 +1482,7 @@ static struct attribute *nhmex_uncore_mbox_formats_attr[] = { &format_attr_flag_mode.attr, &format_attr_inc_sel.attr, &format_attr_set_flag_sel.attr, - &format_attr_filter_cfg.attr, + &format_attr_filter_cfg_en.attr, &format_attr_filter_match.attr, &format_attr_filter_mask.attr, &format_attr_dsp.attr, @@ -1482,6 +1506,12 @@ static struct uncore_event_desc nhmex_uncore_mbox_events[] = { { /* end: all zeroes */ }, }; +static struct uncore_event_desc wsmex_uncore_mbox_events[] = { + INTEL_UNCORE_EVENT_DESC(bbox_cmds_read, "inc_sel=0xd,fvc=0x5000"), + INTEL_UNCORE_EVENT_DESC(bbox_cmds_write, "inc_sel=0xd,fvc=0x5040"), + { /* end: all zeroes */ }, +}; + static struct intel_uncore_ops nhmex_uncore_mbox_ops = { NHMEX_UNCORE_OPS_COMMON_INIT(), .enable_event = nhmex_mbox_msr_enable_event, @@ -1513,7 +1543,7 @@ void nhmex_rbox_alter_er(struct intel_uncore_box *box, struct perf_event *event) struct hw_perf_event_extra *reg1 = &hwc->extra_reg; int port; - /* adjust the main event selector */ + /* adjust the main event selector and extra register index */ if (reg1->idx % 2) { reg1->idx--; hwc->config -= 1 << NHMEX_R_PMON_CTL_EV_SEL_SHIFT; @@ -1522,29 +1552,17 @@ void nhmex_rbox_alter_er(struct intel_uncore_box *box, struct perf_event *event) hwc->config += 1 << NHMEX_R_PMON_CTL_EV_SEL_SHIFT; } - /* adjust address or config of extra register */ + /* adjust extra register config */ port = reg1->idx / 6 + box->pmu->pmu_idx * 4; switch (reg1->idx % 6) { - case 0: - reg1->reg = NHMEX_R_MSR_PORTN_IPERF_CFG0(port); - break; - case 1: - reg1->reg = NHMEX_R_MSR_PORTN_IPERF_CFG1(port); - break; case 2: - /* the 8~15 bits to the 0~7 bits */ + /* shift the 8~15 bits to the 0~7 bits */ reg1->config >>= 8; break; case 3: - /* the 0~7 bits to the 8~15 bits */ + /* shift the 0~7 bits to the 8~15 bits */ reg1->config <<= 8; break; - case 4: - reg1->reg = NHMEX_R_MSR_PORTN_XBR_SET1_MM_CFG(port); - break; - case 5: - reg1->reg = NHMEX_R_MSR_PORTN_XBR_SET2_MM_CFG(port); - break; }; } @@ -1671,7 +1689,7 @@ static int nhmex_rbox_hw_config(struct intel_uncore_box *box, struct perf_event struct hw_perf_event *hwc = &event->hw; struct hw_perf_event_extra *reg1 = &event->hw.extra_reg; struct hw_perf_event_extra *reg2 = &event->hw.branch_reg; - int port, idx; + int idx; idx = (event->hw.config & NHMEX_R_PMON_CTL_EV_SEL_MASK) >> NHMEX_R_PMON_CTL_EV_SEL_SHIFT; @@ -1681,27 +1699,11 @@ static int nhmex_rbox_hw_config(struct intel_uncore_box *box, struct perf_event reg1->idx = idx; reg1->config = event->attr.config1; - port = idx / 6 + box->pmu->pmu_idx * 4; - idx %= 6; - switch (idx) { - case 0: - reg1->reg = NHMEX_R_MSR_PORTN_IPERF_CFG0(port); - break; - case 1: - reg1->reg = NHMEX_R_MSR_PORTN_IPERF_CFG1(port); - break; - case 2: - case 3: - reg1->reg = NHMEX_R_MSR_PORTN_QLX_CFG(port); - break; + switch (idx % 6) { case 4: case 5: - if (idx == 4) - reg1->reg = NHMEX_R_MSR_PORTN_XBR_SET1_MM_CFG(port); - else - reg1->reg = NHMEX_R_MSR_PORTN_XBR_SET2_MM_CFG(port); - reg2->config = event->attr.config2; hwc->config |= event->attr.config & (~0ULL << 32); + reg2->config = event->attr.config2; break; }; return 0; @@ -1727,28 +1729,34 @@ static void nhmex_rbox_msr_enable_event(struct intel_uncore_box *box, struct per struct hw_perf_event *hwc = &event->hw; struct hw_perf_event_extra *reg1 = &hwc->extra_reg; struct hw_perf_event_extra *reg2 = &hwc->branch_reg; - int idx, er_idx; + int idx, port; - idx = reg1->idx % 6; - er_idx = idx; - if (er_idx > 2) - er_idx--; - er_idx += (reg1->idx / 6) * 5; + idx = reg1->idx; + port = idx / 6 + box->pmu->pmu_idx * 4; - switch (idx) { + switch (idx % 6) { case 0: + wrmsrl(NHMEX_R_MSR_PORTN_IPERF_CFG0(port), reg1->config); + break; case 1: - wrmsrl(reg1->reg, reg1->config); + wrmsrl(NHMEX_R_MSR_PORTN_IPERF_CFG1(port), reg1->config); break; case 2: case 3: - wrmsrl(reg1->reg, nhmex_rbox_shared_reg_config(box, er_idx)); + wrmsrl(NHMEX_R_MSR_PORTN_QLX_CFG(port), + nhmex_rbox_shared_reg_config(box, 2 + (idx / 6) * 5)); break; case 4: + wrmsrl(NHMEX_R_MSR_PORTN_XBR_SET1_MM_CFG(port), + hwc->config >> 32); + wrmsrl(NHMEX_R_MSR_PORTN_XBR_SET1_MATCH(port), reg1->config); + wrmsrl(NHMEX_R_MSR_PORTN_XBR_SET1_MASK(port), reg2->config); + break; case 5: - wrmsrl(reg1->reg, reg1->config); - wrmsrl(reg1->reg + 1, hwc->config >> 32); - wrmsrl(reg1->reg + 2, reg2->config); + wrmsrl(NHMEX_R_MSR_PORTN_XBR_SET2_MM_CFG(port), + hwc->config >> 32); + wrmsrl(NHMEX_R_MSR_PORTN_XBR_SET2_MATCH(port), reg1->config); + wrmsrl(NHMEX_R_MSR_PORTN_XBR_SET2_MASK(port), reg2->config); break; }; @@ -1756,8 +1764,8 @@ static void nhmex_rbox_msr_enable_event(struct intel_uncore_box *box, struct per (hwc->config & NHMEX_R_PMON_CTL_EV_SEL_MASK)); } -DEFINE_UNCORE_FORMAT_ATTR(xbr_match, xbr_match, "config:32-63"); -DEFINE_UNCORE_FORMAT_ATTR(xbr_mm_cfg, xbr_mm_cfg, "config1:0-63"); +DEFINE_UNCORE_FORMAT_ATTR(xbr_mm_cfg, xbr_mm_cfg, "config:32-63"); +DEFINE_UNCORE_FORMAT_ATTR(xbr_match, xbr_match, "config1:0-63"); DEFINE_UNCORE_FORMAT_ATTR(xbr_mask, xbr_mask, "config2:0-63"); DEFINE_UNCORE_FORMAT_ATTR(qlx_cfg, qlx_cfg, "config1:0-15"); DEFINE_UNCORE_FORMAT_ATTR(iperf_cfg, iperf_cfg, "config1:0-31"); @@ -2303,6 +2311,7 @@ int uncore_pmu_event_init(struct perf_event *event) event->hw.idx = -1; event->hw.last_tag = ~0ULL; event->hw.extra_reg.idx = EXTRA_REG_NONE; + event->hw.branch_reg.idx = EXTRA_REG_NONE; if (event->attr.config == UNCORE_FIXED_EVENT) { /* no fixed counter */ @@ -2373,7 +2382,7 @@ static void __init uncore_type_exit(struct intel_uncore_type *type) type->attr_groups[1] = NULL; } -static void uncore_types_exit(struct intel_uncore_type **types) +static void __init uncore_types_exit(struct intel_uncore_type **types) { int i; for (i = 0; types[i]; i++) @@ -2814,7 +2823,13 @@ static int __init uncore_cpu_init(void) snbep_uncore_cbox.num_boxes = max_cores; msr_uncores = snbep_msr_uncores; break; - case 46: + case 46: /* Nehalem-EX */ + uncore_nhmex = true; + case 47: /* Westmere-EX aka. Xeon E7 */ + if (!uncore_nhmex) + nhmex_uncore_mbox.event_descs = wsmex_uncore_mbox_events; + if (nhmex_uncore_cbox.num_boxes > max_cores) + nhmex_uncore_cbox.num_boxes = max_cores; msr_uncores = nhmex_msr_uncores; break; default: diff --git a/arch/x86/kernel/cpu/perf_event_intel_uncore.h b/arch/x86/kernel/cpu/perf_event_intel_uncore.h index c9e5dc56630a..5b81c1856aac 100644 --- a/arch/x86/kernel/cpu/perf_event_intel_uncore.h +++ b/arch/x86/kernel/cpu/perf_event_intel_uncore.h @@ -230,6 +230,7 @@ #define NHMEX_S1_MSR_MASK 0xe5a #define NHMEX_S_PMON_MM_CFG_EN (0x1ULL << 63) +#define NHMEX_S_EVENT_TO_R_PROG_EV 0 /* NHM-EX Mbox */ #define NHMEX_M0_MSR_GLOBAL_CTL 0xca0 @@ -275,18 +276,12 @@ NHMEX_M_PMON_CTL_INC_SEL_MASK | \ NHMEX_M_PMON_CTL_SET_FLAG_SEL_MASK) - -#define NHMEX_M_PMON_ZDP_CTL_FVC_FVID_MASK 0x1f -#define NHMEX_M_PMON_ZDP_CTL_FVC_BCMD_MASK (0x7 << 5) -#define NHMEX_M_PMON_ZDP_CTL_FVC_RSP_MASK (0x7 << 8) -#define NHMEX_M_PMON_ZDP_CTL_FVC_PBOX_INIT_ERR (1 << 23) -#define NHMEX_M_PMON_ZDP_CTL_FVC_MASK \ - (NHMEX_M_PMON_ZDP_CTL_FVC_FVID_MASK | \ - NHMEX_M_PMON_ZDP_CTL_FVC_BCMD_MASK | \ - NHMEX_M_PMON_ZDP_CTL_FVC_RSP_MASK | \ - NHMEX_M_PMON_ZDP_CTL_FVC_PBOX_INIT_ERR) +#define NHMEX_M_PMON_ZDP_CTL_FVC_MASK (((1 << 11) - 1) | (1 << 23)) #define NHMEX_M_PMON_ZDP_CTL_FVC_EVENT_MASK(n) (0x7 << (11 + 3 * (n))) +#define WSMEX_M_PMON_ZDP_CTL_FVC_MASK (((1 << 12) - 1) | (1 << 24)) +#define WSMEX_M_PMON_ZDP_CTL_FVC_EVENT_MASK(n) (0x7 << (12 + 3 * (n))) + /* * use the 9~13 bits to select event If the 7th bit is not set, * otherwise use the 19~21 bits to select event. @@ -368,6 +363,7 @@ struct intel_uncore_type { unsigned num_shared_regs:8; unsigned single_fixed:1; unsigned pair_ctr_ctl:1; + unsigned *msr_offsets; struct event_constraint unconstrainted; struct event_constraint *constraints; struct intel_uncore_pmu *pmus; @@ -485,29 +481,31 @@ unsigned uncore_pci_perf_ctr(struct intel_uncore_box *box, int idx) return idx * 8 + box->pmu->type->perf_ctr; } -static inline -unsigned uncore_msr_box_ctl(struct intel_uncore_box *box) +static inline unsigned uncore_msr_box_offset(struct intel_uncore_box *box) +{ + struct intel_uncore_pmu *pmu = box->pmu; + return pmu->type->msr_offsets ? + pmu->type->msr_offsets[pmu->pmu_idx] : + pmu->type->msr_offset * pmu->pmu_idx; +} + +static inline unsigned uncore_msr_box_ctl(struct intel_uncore_box *box) { if (!box->pmu->type->box_ctl) return 0; - return box->pmu->type->box_ctl + - box->pmu->type->msr_offset * box->pmu->pmu_idx; + return box->pmu->type->box_ctl + uncore_msr_box_offset(box); } -static inline -unsigned uncore_msr_fixed_ctl(struct intel_uncore_box *box) +static inline unsigned uncore_msr_fixed_ctl(struct intel_uncore_box *box) { if (!box->pmu->type->fixed_ctl) return 0; - return box->pmu->type->fixed_ctl + - box->pmu->type->msr_offset * box->pmu->pmu_idx; + return box->pmu->type->fixed_ctl + uncore_msr_box_offset(box); } -static inline -unsigned uncore_msr_fixed_ctr(struct intel_uncore_box *box) +static inline unsigned uncore_msr_fixed_ctr(struct intel_uncore_box *box) { - return box->pmu->type->fixed_ctr + - box->pmu->type->msr_offset * box->pmu->pmu_idx; + return box->pmu->type->fixed_ctr + uncore_msr_box_offset(box); } static inline @@ -515,7 +513,7 @@ unsigned uncore_msr_event_ctl(struct intel_uncore_box *box, int idx) { return box->pmu->type->event_ctl + (box->pmu->type->pair_ctr_ctl ? 2 * idx : idx) + - box->pmu->type->msr_offset * box->pmu->pmu_idx; + uncore_msr_box_offset(box); } static inline @@ -523,7 +521,7 @@ unsigned uncore_msr_perf_ctr(struct intel_uncore_box *box, int idx) { return box->pmu->type->perf_ctr + (box->pmu->type->pair_ctr_ctl ? 2 * idx : idx) + - box->pmu->type->msr_offset * box->pmu->pmu_idx; + uncore_msr_box_offset(box); } static inline diff --git a/arch/x86/mm/hugetlbpage.c b/arch/x86/mm/hugetlbpage.c index f6679a7fb8ca..b91e48512425 100644 --- a/arch/x86/mm/hugetlbpage.c +++ b/arch/x86/mm/hugetlbpage.c @@ -56,9 +56,16 @@ static int vma_shareable(struct vm_area_struct *vma, unsigned long addr) } /* - * search for a shareable pmd page for hugetlb. + * Search for a shareable pmd page for hugetlb. In any case calls pmd_alloc() + * and returns the corresponding pte. While this is not necessary for the + * !shared pmd case because we can allocate the pmd later as well, it makes the + * code much cleaner. pmd allocation is essential for the shared case because + * pud has to be populated inside the same i_mmap_mutex section - otherwise + * racing tasks could either miss the sharing (see huge_pte_offset) or select a + * bad pmd for sharing. */ -static void huge_pmd_share(struct mm_struct *mm, unsigned long addr, pud_t *pud) +static pte_t * +huge_pmd_share(struct mm_struct *mm, unsigned long addr, pud_t *pud) { struct vm_area_struct *vma = find_vma(mm, addr); struct address_space *mapping = vma->vm_file->f_mapping; @@ -68,9 +75,10 @@ static void huge_pmd_share(struct mm_struct *mm, unsigned long addr, pud_t *pud) struct vm_area_struct *svma; unsigned long saddr; pte_t *spte = NULL; + pte_t *pte; if (!vma_shareable(vma, addr)) - return; + return (pte_t *)pmd_alloc(mm, pud, addr); mutex_lock(&mapping->i_mmap_mutex); vma_prio_tree_foreach(svma, &iter, &mapping->i_mmap, idx, idx) { @@ -97,7 +105,9 @@ static void huge_pmd_share(struct mm_struct *mm, unsigned long addr, pud_t *pud) put_page(virt_to_page(spte)); spin_unlock(&mm->page_table_lock); out: + pte = (pte_t *)pmd_alloc(mm, pud, addr); mutex_unlock(&mapping->i_mmap_mutex); + return pte; } /* @@ -142,8 +152,9 @@ pte_t *huge_pte_alloc(struct mm_struct *mm, } else { BUG_ON(sz != PMD_SIZE); if (pud_none(*pud)) - huge_pmd_share(mm, addr, pud); - pte = (pte_t *) pmd_alloc(mm, pud, addr); + pte = huge_pmd_share(mm, addr, pud); + else + pte = (pte_t *)pmd_alloc(mm, pud, addr); } } BUG_ON(pte && !pte_none(*pte) && !pte_huge(*pte)); diff --git a/arch/x86/mm/pageattr.c b/arch/x86/mm/pageattr.c index 931930a96160..a718e0d23503 100644 --- a/arch/x86/mm/pageattr.c +++ b/arch/x86/mm/pageattr.c @@ -919,13 +919,11 @@ static int change_page_attr_set_clr(unsigned long *addr, int numpages, /* * On success we use clflush, when the CPU supports it to - * avoid the wbindv. If the CPU does not support it, in the - * error case, and during early boot (for EFI) we fall back - * to cpa_flush_all (which uses wbinvd): + * avoid the wbindv. If the CPU does not support it and in the + * error case we fall back to cpa_flush_all (which uses + * wbindv): */ - if (early_boot_irqs_disabled) - __cpa_flush_all((void *)(long)cache); - else if (!ret && cpu_has_clflush) { + if (!ret && cpu_has_clflush) { if (cpa.flags & (CPA_PAGES_ARRAY | CPA_ARRAY)) { cpa_flush_array(addr, numpages, cache, cpa.flags, pages); diff --git a/arch/x86/platform/efi/efi.c b/arch/x86/platform/efi/efi.c index 2dc29f51e75a..92660edaa1e7 100644 --- a/arch/x86/platform/efi/efi.c +++ b/arch/x86/platform/efi/efi.c @@ -234,7 +234,22 @@ static efi_status_t __init phys_efi_set_virtual_address_map( return status; } -static int efi_set_rtc_mmss(unsigned long nowtime) +static efi_status_t __init phys_efi_get_time(efi_time_t *tm, + efi_time_cap_t *tc) +{ + unsigned long flags; + efi_status_t status; + + spin_lock_irqsave(&rtc_lock, flags); + efi_call_phys_prelog(); + status = efi_call_phys2(efi_phys.get_time, virt_to_phys(tm), + virt_to_phys(tc)); + efi_call_phys_epilog(); + spin_unlock_irqrestore(&rtc_lock, flags); + return status; +} + +int efi_set_rtc_mmss(unsigned long nowtime) { int real_seconds, real_minutes; efi_status_t status; @@ -263,7 +278,7 @@ static int efi_set_rtc_mmss(unsigned long nowtime) return 0; } -static unsigned long efi_get_time(void) +unsigned long efi_get_time(void) { efi_status_t status; efi_time_t eft; @@ -606,13 +621,18 @@ static int __init efi_runtime_init(void) } /* * We will only need *early* access to the following - * EFI runtime service before set_virtual_address_map + * two EFI runtime services before set_virtual_address_map * is invoked. */ + efi_phys.get_time = (efi_get_time_t *)runtime->get_time; efi_phys.set_virtual_address_map = (efi_set_virtual_address_map_t *) runtime->set_virtual_address_map; - + /* + * Make efi_get_time can be called before entering + * virtual mode. + */ + efi.get_time = phys_efi_get_time; early_iounmap(runtime, sizeof(efi_runtime_services_t)); return 0; @@ -700,10 +720,12 @@ void __init efi_init(void) efi_enabled = 0; return; } +#ifdef CONFIG_X86_32 if (efi_native) { x86_platform.get_wallclock = efi_get_time; x86_platform.set_wallclock = efi_set_rtc_mmss; } +#endif #if EFI_DEBUG print_efi_memmap(); diff --git a/arch/x86/realmode/rm/Makefile b/arch/x86/realmode/rm/Makefile index b2d534cab25f..88692871823f 100644 --- a/arch/x86/realmode/rm/Makefile +++ b/arch/x86/realmode/rm/Makefile @@ -72,7 +72,7 @@ KBUILD_CFLAGS := $(LINUXINCLUDE) -m32 -g -Os -D_SETUP -D__KERNEL__ -D_WAKEUP \ -Wall -Wstrict-prototypes \ -march=i386 -mregparm=3 \ -include $(srctree)/$(src)/../../boot/code16gcc.h \ - -fno-strict-aliasing -fomit-frame-pointer \ + -fno-strict-aliasing -fomit-frame-pointer -fno-pic \ $(call cc-option, -ffreestanding) \ $(call cc-option, -fno-toplevel-reorder,\ $(call cc-option, -fno-unit-at-a-time)) \ diff --git a/arch/x86/syscalls/syscall_64.tbl b/arch/x86/syscalls/syscall_64.tbl index 29aed7ac2c02..a582bfed95bb 100644 --- a/arch/x86/syscalls/syscall_64.tbl +++ b/arch/x86/syscalls/syscall_64.tbl @@ -60,8 +60,8 @@ 51 common getsockname sys_getsockname 52 common getpeername sys_getpeername 53 common socketpair sys_socketpair -54 common setsockopt sys_setsockopt -55 common getsockopt sys_getsockopt +54 64 setsockopt sys_setsockopt +55 64 getsockopt sys_getsockopt 56 common clone stub_clone 57 common fork stub_fork 58 common vfork stub_vfork @@ -353,3 +353,5 @@ 538 x32 sendmmsg compat_sys_sendmmsg 539 x32 process_vm_readv compat_sys_process_vm_readv 540 x32 process_vm_writev compat_sys_process_vm_writev +541 x32 setsockopt compat_sys_setsockopt +542 x32 getsockopt compat_sys_getsockopt |
